www.elsevier.com/locate/ydbio

Developmental Biology 272 (2004) 118–133

Neural cells in the esophagus respond to glial cell line-derived

neurotrophic factor and neurturin, and are RET-dependent

Hui Yan,a Annette J. Bergner,a Hideki Enomoto,b,c Jeffrey Milbrandt,c

Donald F. Newgreen,d and Heather M. Younga,*

aDepartment of Anatomy and Cell Biology, University of Melbourne, Parkville, 3010 Victoria, AustraliabLaboratory for Neuronal Differentiation and Regeneration, RIKEN Center for Developmental Biology, Chuo, Kobe, Hyogo 650-0047, Japan

cDepartments of Pathology and Internal Medicine, Washington University School of Medicine, St. Louis, MO 63110, USAdMurdoch Childrens Research Institute, Royal Childrens Hospital, Parkville, 3052 Victoria, Australia

Received for publication 12 February 2004, revised 12 April 2004, accepted 12 April 2004

Available online 24 May 2004

Abstract

Glial cell line-derived neurotrophic factor (GDNF) is expressed in the gastrointestinal tract of the developing mouse and appears to

play an important role in the migration of enteric neuron precursors into and along the small and large intestines. Two other GDNF family

members, neurturin and artemin, are also expressed in the developing gut although artemin is only expressed in the esophagus. We

examined the effects of GDNF, neurturin, and artemin on neural crest cell migration and neurite outgrowth in explants of mouse

esophagus, midgut, and hindgut. Both GDNF and neurturin induced neural crest cell migration and neurite outgrowth in all regions

examined. In the esophagus, the effect of GDNF on migration and neurite outgrowth declined with age between E11.5 and E14.5, but

neurturin still had a strong neurite outgrowth effect at E14.5. Artemin did not promote neural migration or neurite outgrowth in any region

investigated. The effects of GDNF family ligands are mediated by the Ret tyrosine kinase. We examined the density of neurons in the

esophagus of Ret�/� mice, which lack neurons in the small and large intestines. The density of esophageal neurons in Ret�/� mice was

only about 4% of the density of esophageal neurons in Ret+/� and Ret+/+ mice. These results show that GDNF and neurturin promote

migration and neurite outgrowth of crest-derived cells in the esophagus as well as the intestine. Moreover, like intestinal neurons, the

development of esophageal neurons is largely Ret-dependent.

D 2004 Elsevier Inc. All rights reserved.

Keywords: GDNF; Neurturin; Cell migration; Neurite outgrowth; Ret�/�; Esophagus; Gastrointestinal tract

Introduction ligand binding molecule, GFRa1, and a receptor tyrosine

The enteric nervous system is an extensive system of

neurons and glial cells within the wall of the gastrointestinal

tract. Most enteric neurons arise from neural crest cells that

emigrate from the vagal level of the neural axis (post-otic

hindbrain adjacent to somites 1–7) (Burns and Le Douarin,

1998; Le Douarin and Teillet, 1973; Yntema and Hammond,

1954).

The glial cell line-derived neurotrophic factor (GDNF)

signaling pathway is essential for the development of the

enteric nervous system in most regions of the gastrointes-

tinal tract. GDNF acts at a receptor complex consisting of a

0012-1606/$ - see front matter D 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.ydbio.2004.04.025

* Corresponding author. Fax: +61-613-9347-5219.

E-mail address: h.young@unimelb.edu.au (H.M. Young).

kinase, Ret (Airaksinen et al., 1999). Mice lacking GDNF,

GFRa1, or Ret lack enteric neurons in the small and large

intestines (Cacalano et al., 1998; Enomoto et al., 1998;

Moore et al., 1996; Pichel et al., 1996; Sanchez et al., 1996;

Schuchardt et al., 1994; Tomac et al., 2000). GDNF appears

to play multiple roles in enteric neuron development,

including survival, proliferation, and differentiation (Chala-

zonitis et al., 1998; Focke et al., 2001, 2003; Gianino et al.,

2003; Hearn et al., 1998; Shen et al., 2002; Taraviras et al.,

1999; Worley et al., 2000). In addition, GDNF is chemo-

attractive to enteric neural crest-derived cells, and appears to

play an important role in promoting the migration of vagal

neural crest cells into and along the gastrointestinal tract

(Barlow et al., 2003; Iwashita et al., 2003; Kruger et al.,

2003; Natarajan et al., 2002; Young et al., 2001). Despite

the absence of enteric neurons in the small and large

Fig. 1. Diagram of the gastrointestinal tract from an E11.5 mouse showing

the location of neural crest-derived cells (shown in grey; see Young and

H. Yan et al. / Developmental Biology 272 (2004) 118–133 119

intestines of mice with null mutations in Gdnf, Gfra1, orRet, enteric neurons are reported to be present in the

esophagus of Ret�/� mice, and, therefore, it has been

proposed that the development of neurons in the esophagus

is Ret-independent (Durbec et al., 1996). Thus, enteric

neuron precursors in different parts of the gastrointestinal

tract may differ in their requirements for GDNF family

members.

In addition to GDNF, two other members of the GDNF

family of ligands are expressed in the developing gastroin-

testinal tract, neurturin and artemin. The effects of neurturin

are mediated by a receptor complex consisting of its

preferred co-receptor, GFRa2, and Ret, and the receptor

complex for artemin consists of GFRa3 and Ret, although

neurturin and artemin may also bind weakly to GFRa1

(Airaksinen and Saarma, 2002). Neurturin is expressed

throughout the developing gastrointestinal tract (Golden et

al., 1999; Widenfalk et al., 1997; Xian et al., 1999), whereas

artemin is only expressed in the esophagus (Enomoto et al.,

2001). Neurturin promotes the proliferation and differenti-

ation of enteric neural crest-derived cells in vitro (Heuck-

eroth et al., 1998; Taraviras et al., 1999). Nonetheless, mice

with null mutations in neurturin or GFRa2 have neurons

present in normal numbers throughout the gastrointestinal

tract, but show a decreased density of excitatory nerve fibres

in the circular muscle (Gianino et al., 2003; Heuckeroth et

al., 1999; Rossi et al., 1999, 2003). Studies of mice with a

null mutation in artemin or GFRa3, the binding molecule

for artemin, have not reported any defects in the enteric

nervous system (Honma et al., 2002; Nishino et al., 1999).

The effects of neurturin and artemin on enteric neural crest

cell-derived migration have not previously been examined.

In this study, we examined the effect of GDNF, neurturin,

and artemin on neural migration and neurite outgrowth from

neural precursors in different regions of the E10.5–E14.5

gastrointestinal tract (esophagus, midgut, and hindgut). Our

results show that both GDNF and neurturin induce neural

migration and neurite outgrowth in the esophagus as well as

the intestine, but the effect of GDNF declines with age.

Although artemin is expressed in the esophagus, artemin did

not induce migration or neurite outgrowth in the esophagus

or any gastrointestinal region examined. In addition, al-

though mice lacking Ret have been reported to possess

enteric neurons in the esophagus, the density of enteric

neurons in the esophagus of Ret�/� mice has not been

previously been quantified. We show that the density of

neurons in the esophagus of Ret�/� mice is dramatically

(about 25 times) lower than that in Ret+/� or Ret+/+ mice,

suggesting that the development of enteric neurons in all

regions of the gastrointestinal tract is largely Ret-dependent.

Newgreen, 2001). The regions labelled ‘‘a’’ (esophagus), ‘‘b’’ (midgut), and

‘‘c’’ (hindgut) were removed, suspended between a ‘‘V’’ cut into a piece of

filter paper and grown in tissue culture medium for 4 days as described

previously (Hearn et al., 1999). At the time of explantation, neural crest-

derived cells (grey) were present throughout the explants of esophagus and

midgut, whereas in the hindgut, they were only present at the rostral end of

the explants (lower panels).

Materials and methods

Embryonic BALB/c mice were used for all organ culture

experiments. The day at which a vaginal plug was found

was designated E0.5. Pregnant mothers were killed by

cervical dislocation, and the embryos removed using aseptic

conditions.

Suspension (catenary) culture

Segments of esophagus, midgut, and hindgut from E11.5

mice, esophagus from E10.5, and midgut from E12.5 mice

were dissected. The entire esophagus, which was 1.2–1.5

mm long, was dissected and explanted. Explants of midgut

were from the caudal region, rostral to the caecum, and

include regions that will develop into the ileum and jeju-

num. A ‘‘V’’ was cut into a 3 � 3-mm piece of Millipore

filter paper using a scalpel, and an explant was suspended

across the ‘‘V’’ as described previously (Hearn et al., 1999;

Fig. 1). The orientation of the hindgut explant was indicated

by removing the corner of the filter paper adjacent to the

caudal end. The whole preparation, including the filter paper

and the attached explant, was placed in a Terasaki well in 20

Al of tissue culture medium and cultured in a 5% CO2

incubator at 37jC for 4 days. The tissue culture medium

contained DMEM (Trace Scientific Ltd., Noble Park, Vic-

toria, Australia) with 10% fetal bovine serum, 2 mM L-

glutamine, 0.075% sodium bicarbonate, and penicillin/strep-

tomycin sulfate solution. GDNF, artemin, or neurturin (all at

100 ng/ml, PeproTech Inc., Rocky Hill, NJ, USA) was

added to the culture medium of some explants. The culture

medium was changed after 2 days.

Table 1

Primary and secondary antibodies used

Primary antibodies

Host Dilution Source

PGP 9.5 Rabbit 1:2000 The Binding Site,

Birmingham, UK

Neurofilament-M

(145 kDa)

Rabbit 1:1000 Chemicon, Temecula,

CA, USA

Hu Human 1:2000 Fairman et al., 1995

NOS Sheep 1:2000 Norris et al., 1995

Phox2b Rabbit 1:700 Pattyn et al., 1997

GFRa1, GFRa3 Goat 1:25 R&D systems, Inc.,

Minneapolis, MN, USA

Secondary antibodies

Species in which

primary antibodies

were raised

Secondary antibodies

Rabbit Goat anti-rabbit Alexa 488 (1:250, Molecular

Probes, Eugene, OR, USA)

Human Donkey anti-human Texas Red (1:100, Jackson

ImmunoResearch, West Grove, PA, USA)

Sheep or goat Donkey anti-sheep FITC (1:100, Jackson

ImmunoResearch), or Biotinylated donkey

anti-sheep (1:100, Jackson ImmunoResearch)

followed by streptavidin-Cy5

(1:100, Amersham)

tal Biology 272 (2004) 118–133

Slice explants grown on collagen gel

The esophagus and midgut from E11.5, E12.5, and E14.5

mice were dissected, cut into transverse sections about 0.5–1

mm thick, and the slices placed on collagen gels in a 5% CO2

incubator at 37jC for 4 days. The collagen gels were made

by restoring 4 mg/ml acidic collagen solution (Upstate,

Parkville, Victoria, Australia) to normal osmolality with

5� DMEM and normal pH with 200 mM NaOH, on ice.

This solution was diluted to 1 mg/ml collagen with tissue

culture medium. GDNF, artemin, or neurturin was added to

some collagen solutions before gelling to give a final

concentration of 100 ng/ml.

Slice explants grown on filter paper

Transverse slices of esophagus and midgut from E11.5

mice were placed on 2.5 � 2.5 mm squares of filter papers.

Each preparation was placed in a Terasaki well in 20 Al oftissue culture medium with or without 100 ng/ml GDNF,

artemin, or neurturin and cultured in a 5% CO2 incubator at

37jC for 4 days. The tissue culturemediumwas changed after

2 days.

Immunohistochemistry

Explants were fixed in 4% paraformaldehyde in 0.1 M

phosphate buffer for 4–24 hours. Explants grown on filter

paper supports were left attached to the filter paper. After

fixation, the explants were processed for immunohistochem-

istry using the antisera shown in Table 1. Whole-mount

preparations of the gastrointestinal tract from the esophagus

to the rectum from E12.5 and E14.5 mice were fixed for 2–

4 hours in 4% paraformaldehyde in 0.1 M phosphate buffer

or in Zamboni’s fixative for 2–4 hours, and then processed

for immunohistochemistry using GFRa1 and GFRa3 anti-

sera (see Table 1).

Neuron counts within and outside suspended explants of gut

The total number of PGP9.5+ neurons present on the

filter paper supports outside the gut proper was counted,

and the density of PGP9.5+ neurons on the top surface of

the gut explants was determined. To count the total

number of PGP9.5+ cells on the filter paper, images were

taken on a fluorescence microscope using a 5� lens. A

montage was made of the images using CorelDRAW10

and printed, and PGP9.5+ neurons were counted from the

print. In regions where the cells were clumped or over-

lapped, a 40� lens was used to count the neurons directly

under the microscope. The total numbers of PGP9.5+

neurons on the paper supports at both rostral and caudal

ends were counted and added. Within the gut, the density

of PGP9.5+ neurons on the top surface only of the

suspended explants was determined as the neurons along

the bottom and sides of the gut explants could not be

H. Yan et al. / Developmen120

visualized clearly because of the tubular shape of the gut.

The preparations were viewed on Bio-Rad MRC1000 or

1024 confocal microscopes (Bio-Rad, Richmond, CA) and

images of PGP9.5+ neurons in the suspended gut were

taken by using a 20� objective lens. Images were printed

and the number of neurons counted. The area of gut

explant within which the neurons were counted was

measured using Scion image software and then the density

of neurons was calculated.

Enteric neuron density within the esophagus of Ret�/�,

Ret+/�, and Ret+/+ mice

Mice in which cDNA encoding tau-EGFP-myc (TGM)

had been inserted into the first coding exon of the receptor

tyrosine kinase gene, Ret, were used (Enomoto et al., 2001).

Mice heterozygous for the Ret-TGM mutation (RetTGM/+

mice) were mated. The genotype of adult and embryonic

RetTGM mice was determined by PCR using the primers and

conditions reported by Enomoto et al. (2001). The identity

of E18.5 Ret�/� mice was confirmed by processing samples

of duodenum from each embryo for NADPH diaphorase

histochemistry (Ward et al., 1999). Samples of duodenum

were fixed in 4% paraformaldehyde in 0.1M phosphate

buffer for 1 h, washed, incubated in 10 mg h-NADPH,2.5 mg nitroblue tetrazolium, and 20 Al Triton X in 10 ml

0.1 M Tris/HCl (pH 8.0) for 15–20 minutes at 37jC and

then examined to determine if there were any NADPH

diaphorase-stained neurons. E18.5 esophagus was processed

for immunohistochemistry using antibodies to Hu (Table 1).

Fig. 2. Fluorescence micrographs of explants of E11.5 hindgut grown in control conditions (A, C) or in the presence of neurturin (B, D) for 4 days and then

processed for immunohistochemistry using an antibody to the pan-neuronal marker, PGP 9.5. (A, B) The rostral ends of the explants. A small number of

PGP9.5+ neurons (arrows) are present on the filter paper support under control conditions (A). (B) After culturing in the presence of neurturin, many PGP9.5+

neurons (arrows) are present on the filter paper support. (C, D) PGP9.5+ neurons within hindgut explants. The density of neurons is higher in the control

explant (C) than in the explant grown in the presence of neurturin (D). Scale bars: 100 Am (A, B); 50 Am (C, D). Effects of GDNF, artemin, and neurturin on the

total number of PGP9.5+ neurons on the filter paper supports outside of E11.5 hindgut explants (E), and the density of neurons within the explants (F). Data are

means F SEM. In the presence of GDNF or neurturin, both the total number of neurons on the filter paper and the densities of neurons within the suspended

hindgut explants are significantly different from those in control conditions, but explants grown in the presence of artemin do not show a significant difference

from controls in either the number of neurons on the filter paper or the density of neurons within the explant (ANOVA followed by a Tukey test).

H. Yan et al. / Developmental Biology 272 (2004) 118–133 121

H. Yan et al. / Developmental Biology 272 (2004) 118–133122

The density of Hu+ cells on the top surface of each

esophagus was determined using the same technique as that

used to determine the density of neurons in suspension

cultures (see above) except that images were taken on a

conventional fluorescence microscope using a 10� objec-

tive lens. In addition, the proportion of Hu+ neurons that

showed NOS immunostaining was determined in some

E18.5 preparations. The entire gastrointestinal tract from

E11.5 Ret transgenic mice was processed for immunohisto-

Fig. 3. The rostral ends of explants of E11.5 esophagus grown in control conditio

processed for immunohistochemistry using an antibody to the pan-neuronal mar

supports under all three conditions. Scale bars: 100 Am. Effects of GDNF, artemin

outside of E11.5 esophagus explants (D) and on the density of neurons within the

between the four groups in the number of neurons on the filter paper supports,

neurturin was significantly lower than that in controls (ANOVA followed by a Tu

GDNF or artemin were not significantly different from controls.

chemistry using an antibody to Phox2b, to ascertain the

entire enteric neural crest cell population.

Statistical analysis

Student’s two-tailed unpaired t tests or repeated measures

ANOVA followed by Tukey post hoc tests were performed

where appropriate. A probability of less than 0.05 (P <

0.05) was considered significant throughout.

ns (A) or in the presence of GDNF (B) or neurturin (C) for 4 days and then

ker, PGP9.5. There are few PGP9.5+ neurons (arrows) on the filter paper

, and neurturin on the total number of neurons on the filter paper supports

explants (E). Data are means F SEM. There was no significant difference

but the density of neurons within the explants cultured in the presence of

key test). The densities of neurons within explants grown in the presence of

Fig. 4. Effects of GDNF on the total number of PGP9.5+ neurons on the

filter paper supports (A) and density of neurons within explants (B) of

explants of E11.5 and E12.5 midgut. Data are means F SEM. (A) At both

E11.5 and E12.5, the total numbers of neurons on the filter paper supports

of midgut explants grown in the presence of GDNF were significantly

higher than those grown in control conditions (unpaired t test). (B) At

E12.5, the density of neurons within the control midgut explants was

significantly higher than the explants grown in the presence of GDNF

(unpaired t test), but at E11.5, there was no significant difference between

the two groups (unpaired t test).

H. Yan et al. / Developmental Biology 272 (2004) 118–133 123

Results

Neural cell migration response induced by GDNF, artemin,

or neurturin in different regions of embryonic mouse gut

grown in suspension organ culture

Hindgut explants

Explants of caecum and post-caecal hindgut from E11.5

mice were grown in suspension organ culture on filter paper

supports in control culture medium or with added GDNF,

artemin, or neurturin (100 ng/ml). At the time of explanta-

tion, neural crest cells are present in the most rostral end of

each hindgut explant only (caecum and rostral-most post-

caecal hindgut) (Fig. 1; Young and Newgreen, 2001). After

4 days in culture, the explants were fixed and processed for

PGP9.5 immunohistochemistry to reveal neurons. The den-

sity of PGP9.5+ cells within the explants, and the total

number of PGP9.5+ cells on the filter paper supports, were

determined.

Under control conditions, many PGP9.5+ neurons were

present throughout the gut explants (Figs. 2C,F) and only a

small number of neurons was found on the filter paper (Figs.

2A,E), mostly at the rostral end of the explants. Thus, neural

crest cells had migrated through the explants and differen-

tiated into neurons. In the presence of artemin, the density of

neurons within the explants and the number of neurons on

the filter paper supports were not significantly different from

controls (ANOVA; Figs. 2E,F). However, in the presence of

GDNF or neurturin, the density of PGP9.5+ neurons within

the suspended gut explants was significantly lower than that

in control explants (Figs. 2D,F), and the number of PGP9.5+

neurons on the filter paper supports was significantly higher

than in controls (ANOVAs; Figs. 2B,E). These data confirm

a previous study that showed that GDNF in the culture

medium can promote the migration of neural crest cells out

of suspended hindgut explants in vitro (Young et al., 2001).

The results also show that neurturin, but not artemin, has

similar effects to GDNF on the migratory behavior of neural

crest cells in the hindgut.

Esophagus explants

Explants of E11.5 esophagus (rostral foregut) were

grown suspended on filter papers supports. Note that

explants of E11.5 esophagus differ from E11.5 hindgut

explants in that neural crest-derived cells are present

throughout the entire rostro-caudal extent of the explant at

the time of explantation (Fig. 1). After 4 days in culture, the

total number of PGP9.5+ neurons present on the filter paper

supports in the presence of GDNF, artemin, or neurturin was

not significantly different from controls (ANOVA; Fig. 3).

The density of neurons within explants of esophagus grown

in the presence of GDNF or artemin was also not signifi-

cantly different from controls, but the density of neurons

within esophagus explants grown in the presence of neu-

rturin was significantly lower than that in control explants

(ANOVA; Fig. 3E). Suspension cultures of E10.5 esophagus

were also attempted to examine if GDNF family members

could induce migration from younger suspended explants of

esophagus. At E10.5, the esophagus is short and very

fragile, and we were unable to grow them successfully as

suspended explants.

Midgut explants

Segments of E11.5 and E12.5 midgut were grown in

suspension cultures with or without added GDNF. The

colonization of the midgut by neural crest-derived cells

begins around E10.5, and by E11.5, the entire midgut has

been colonized (see Young and Newgreen, 2001). Thus, like

the esophagus from E11.5 mice, these segments of midgut

have neural crest-derived cells throughout the entire seg-

ment at the time of explantation. Data showing the effect of

GDNF on PGP9.5+ neuron number outside of, and within,

suspended explants of E11.5 and E12.5 midgut are shown in

Fig. 4. At both E11.5 and E12.5, the number of neurons on

the filter paper supports of E11.5 and E12.5 midgut explants

grown in the presence of GDNF was significantly greater

than controls (unpaired t test; Fig. 4A). However, the effect

H. Yan et al. / Developmental Biology 272 (2004) 118–133124

of GDNF on the number of neurons on the filter paper

declined dramatically with age. In the presence of GDNF,

the density of neurons within the gut explants was signif-

icantly lower than in the controls at E12.5 (unpaired t test),

but not at E11.5 (unpaired t test; Fig. 4B). This difference

may reflect differing effects of GDNF on proliferation and/

or differentiation at different ages. These data demonstrate

that even when neural crest cells are present throughout

midgut explants at the time of explantation, GDNF is

capable of inducing an increase in the number of neurons

on the filter paper supports, but that the effect of GDNF on

Fig. 5. Inverted fluorescence images of transverse slices of E11.5 esophagus grow

and then processed for immunohistochemistry using antibodies to neurofilament (N

the presence of artemin, there were no or few neurites extending out of the explan

muscle layer of the explant (B, D). In the presence of GDNF and neurturin, many n

nerve cell bodies (arrows) were found outside the esophagus explants (F, G). Mos

bars: 50 Am.

the migratory behaviour of neural crest-derived cells in the

midgut declines with age.

Neural crest-derived cell migration and neurite outgrowth

from transverse slices of esophagus and midgut grown on

collagen gels with or without GDNF, artemin, or neurturin

E11.5 esophagus

Transverse slices of esophagus were grown on collagen

gels, with or without 100 ng/ml GDNF, artemin, or neu-

rturin within the gel. After 4 days in culture, the explants

n on collagen gels with or without GDNF, artemin, or neurturin in the gels

F, to label neurites) and Hu (to label nerve cell bodies). In the controls and

ts (A, C), and most Hu+ neurons (arrows) were present within the external

eurites extended from the explant into the collagen gel (E, H) and many Hu+

t of the nerve cell bodies outside of the explants occurred in clumps. Scale

H. Yan et al. / Developmental Biology 272 (2004) 118–133 125

were immunostained for Hu to reveal neural cell bodies and

for neurofilament to reveal neurites. In control conditions, or

with artemin added to the gels, many Hu+ cells were present

within the explant, and they formed a discrete layer in the

outer part of the mesenchyme (Figs. 5B,D). A small number

of Hu+ cells was found close to, but outside, the explants in

a few specimens (2/12 controls and 1/12 with artemin

present). In control or artemin-treated explants, there were

either no neurites, or only sparse neurites, that extended

away from the explants (Figs. 5A,C). In the presence of

GDNF or neurturin, many Hu+ cells were observed on the

gel outside of the explants (Figs. 5F,G). The Hu+ cells

outside the explants were mainly found in clumps (Figs.

5F,G). Most esophagus explants grown on gels containing

GDNF or neurturin also possessed many bundles of neuro-

filament+ neurites that extended away from the explants (8/

12 with GDNF present in the gel, 10/12 with neurturin;

Figs. 5E,H). The clumps of Hu+ cells were usually found

associated with bundles of neurites. Hence, in contrast to

suspended explants of E11.5 esophagus, GDNF and neu-

rturin induced a migratory and neurite outgrowth response

Fig. 6. (A–D) Fluorescence images of transverse slices of E14.5 esophagus grown

immunohistochemistry using antibodies to neurofilament (NF) and Hu. In the prese

explants and rare Hu+ neurons outside of the explants. However, in the presence

explants (D) but few Hu+ cell bodies were present outside of the explant (C). (E)

processed for immunohistochemistry using antisera to Phox2b (red, to label all ne

immunostaining, predominantly on their cell membranes (arrows). Scale bars A–

from transverse slices of E11.5 esophagus grown on colla-

gen gels.

E12.5 and E14.5 esophagus

E12.5 (n = 12) and E14.5 (n = 6) esophageal slices grown

in the presence of GDNF did not show a migratory response

(i.e., no Hu+ cells were present outside of the explants). A

neurite outgrowth response to GDNF was seen in only 42%

(5/12) E12.5 explants and 33% (2/6) of E14.5 explants

exposed to GDNF, but in these explants, only short neuro-

filament+ neurites extended from the explants (Fig. 6B).

However, strong neurite outgrowth responses were induced

by neurturin from the E12.5 (n = 6) and E14.5 (n = 6)

esophagus explants (Fig. 6D), although Hu+ cell bodies were

only rarely (1/6) observed outside the explants (Fig. 6C).

E11.5 midgut

Transverse sections of E11.5 midgut were grown on

collagen gels under control conditions or with GDNF,

neurturin, or artemin added to the gel. In control and

artemin-treated explants, Hu+ cells were observed within

on collagen gels with GDNF or neurturin in the gels and then processed for

nce of GDNF (A, B), there were few neurites extending from the esophagus

of neurturin, many neurites extended onto the collagen gel outside of the

Whole-mount (uncultured) preparation of esophagus from an E14.5 mouse

ural crest cells) and GFRa1 (green). All Phox2b+ cells show strong GFRa1

D: 100 Am, E: 25 Am.

H. Yan et al. / Developmental Bi126

all of the explants, and a small number of Hu+ cells was

observed outside of the explants in some experiments (4/12

of control and 3/12 of artemin-treated) (Figs. 7B,D). Sparse

neurites extended from some of the explants in 7/12 of

control (Fig. 7A) and 3/12 of artemin-treated (Fig. 7C). In

contrast, extensive neurite outgrowth was observed extend-

ing from all midgut slices grown on collagen gels containing

GDNF (n = 12) or neurturin (n = 12) (Figs. 7G,E), and some

of the neurites extended for over 1 mm away from the

explants. In the presence of GDNF or neurturin, many Hu+

cells were observed both within the midgut slice explants

and in clumps associated with neurites outside of the

explants (Figs. 7H,F).

Fig. 7. Inverted fluorescence images of transverse slices of E11.5 midgut grown o

then processed for immunohistochemistry using antibodies to neurofilament (NF) a

there were few neurites extending from the midgut explants and few Hu+ nerve

(arrows) were found outside of the midgut explants in the presence of GDNF (G, H

occurred in clumps. Scale bars: 100 Am.

E12.5 and E14.5 midgut

In the presence of GDNF or neurturin, Hu+ cells were

present outside of the midgut explants and neurites extended

from the explants onto the collagen gel at both E12.5 (n =

12) and E14.5 (n = 6) (data not shown).

Neural crest-derived cell migration and neurite outgrowth

from transverse slices of E11.5 esophagus and midgut

grown on filter paper with or without GDNF, artemin, or

neurturin present in the culture medium

The experiments in which suspended tubes of E11.5

esophagus were grown on filter paper supports suggested

ology 272 (2004) 118–133

n collagen gels with or without GDNF, artemin, or neurturin in the gels and

nd Hu. Under control conditions (A, B) or in the presence of artemin (C, D),

cell bodies outside of the explants. However, many neurites and neurons

) and neurturin (E, F). Most of the nerve cell bodies outside of the explants

H. Yan et al. / Developmental Biology 272 (2004) 118–133 127

that GDNF and neurturin do not promote migration of

esophageal neural cells, whereas the experiments in which

transverse slices of E11.5 esophagus were grown on

collagen gels containing GDNF or neurturin suggested

that both these factors induce migration of neural cells

from the esophagus. To determine whether the differences

were due to differences in the size and shape of the

explants (whole tubes versus transverse slices) or differ-

ence in the substrate (filter paper versus collagen gel),

transverse slices of E11.5 esophagus and midgut were

attached to small squares of filter paper and grown in

culture medium for 4 days.

Esophagus

Similar results were found in control conditions and in

the presence of artemin. Many Hu+ cells were present

within the outer mesenchyme of the esophagus slice

explants, but there were few or no Hu+ cells outside of

the explants (Figs. 8A,B) and few or no neurites extend-

ing from the explant onto the filter paper. In the presence

of GDNF and neurturin, some neurites extended from the

explants onto the filter paper, but the number of neurites

appeared to be less than that observed on collagen gels.

There were very few Hu+ cells present within the

explants, but many Hu+ cells outside of the explants in

the GDNF and neurturin-treated explants (Figs. 8C,D).

However, in contrast to the slice explants grown on

collagen gel, many of the Hu+ cells outside of the

explants were not found in clumps (Figs. 8C,D). Hence,

the migratory response induced by GDNF and neurturin

from transverse slices of E11.5 esophagus onto a filter

paper substrate is similar to that seen in collagen gels.

Midgut

Slices of midgut grown on filter paper responded to

GDNF family members in a similar manner to that observed

when they were grown on collagen gels (data not shown). In

control conditions, many Hu+ cells were present within the

wall of the midgut slices, but no or few neurites extended

from the explants onto the filter paper. Some control

explants had a small number of Hu+ cells outside, but close

to, the explant. In the presence of artemin, a few neurites

(<10) were observed extending from some explants (3/6)

and a small number of Hu+ cells were associated with these

neurites. Many Hu+ cells were present within the explant

wall of artemin-treated explants. In the presence of GDNF

Fig. 8. Fluorescence images of transverse slices of E11.5 esophagus grown

on filter paper in control condition (A), or with artemin (B), GDNF (C), or

neurturin (D) added to the culture medium for 4 days, and then processed

for immunohistochemistry using an antibody to Hu to label nerve cell

bodies. Under control conditions (A) or in the presence of artemin (B), Hu+

cell bodies were present in the external muscle layer of the slices and few

were present outside of the explants. In the presence of GDNF (C) or

neurturin (D), many Hu+ neurons (arrows) were present on the filter paper,

outside of the explants, and there were very few neurons within the explants

(whose locations are indicated with a dotted line). Scale bars: 50 Am.

H. Yan et al. / Developmental Biology 272 (2004) 118–133128

(n = 6) or neurturin (n = 6), many Hu+ cells were present

both within and outside of midgut slice explants and many

neurites extended from the explant onto the filter paper. The

migrating neurons were closely associated with nerve fibres.

GFRa1 and GFRa3 immunostaining

GFRa1

The experiments described above showed that the

GDNF-induced neurite outgrowth and migratory responses

of neural crest-derived cells in the esophagus declined with

age, whereas the response of midgut neural cells to GDNF

was maintained until at least E14.5. To examine whether the

decline in responsiveness of esophageal neural cells is due

to a down-regulation of GFRa1, we examined the presence

of GFRa1 immunostaining in whole-mount preparations of

esophagus and midgut using an antiserum that has previ-

ously been shown to be specific for GFRa1 (Golden et al.,

2003). At E14.5, neural crest cells (identified by Phox2b

immunostaining) in both the esophagus (Fig. 6E) and

midgut showed strong GFRa1 immunostaining, and there

was no obvious difference in the levels of immunostaining

between cells in the two different gut regions. Thus, the

Fig. 9. Density of Hu+ (A) and NOS+ cells as a percentage of Hu+ cells (B) wi

means F SEM. The density of Hu+ cells and the percentage of Hu+ cells that e

than that in wild-type and Ret+/� mice (ANOVAs). There was no significant diffe

whole mounts of esophagus stained using Hu antisera. Many Hu+ cells are pres

esophagus of Ret� /� mice, very few Hu+ cells were present. Scale bar: 50 Am

decline in responsiveness of cells in the esophagus does not

appear to be due to a down-regulation of GFRa1.

GFRa3

Unlike sympathetic precursors (Honma et al., 2002), we

did not detect any migratory or neurite outgrowth response

to artemin in any gut region at any developmental stage

examined. To determine whether the lack of effect of

artemin on enteric crest-derived cells is due to the lack of

expression of GFRa3, we processed whole-mount prepara-

tions of sympathetic ganglia, midgut, and esophagus from

E12.5 and E14.5 mice for GFRa3 immunostaining. Al-

though GFRa3 immunostaining was observed in sympa-

thetic ganglia, we did not observe any detectable GFRa3

immunostaining in crest-derived cells in either the esopha-

gus or midgut (data not shown).

Comparison of neural crest-derived cell density within the

esophagus of E11.5 and E18.5 Ret+/+, Ret+/�, and Ret�/�

mice

The experiments described above showed that neural

crest-derived cells in the esophagus show similar neural

thin the esophagus of E18.5 Ret+/ + , Ret+/� , and Ret� /� mice. Data are

xpress NOS within the esophagus of Ret� / � mice are significantly lower

rence between the wild-type mice and Ret+/� mice. (C–E) Micrographs of

ent within the esophagus of Ret+/ + (C) and Ret+/� (D) mice. (E) In the

(applies to C–E).

H. Yan et al. / Developmental Biology 272 (2004) 118–133 129

migration responses to GDNF and neurturin to cells in the

midgut and hindgut. A previous study reported that

enteric neurons are present in the esophagus of Ret�/�

mice (Durbec et al., 1996). However, the number of

neurons present in the esophagus of mice lacking mem-

bers of the GDNF-signalling pathway has not been

examined quantitatively.

E18.5 esophagus

We examined the density of all enteric neurons in the

E18.5 esophagus using Hu antisera. The identity of Ret�/�

mice determined by PCR was confirmed by NADPH

diaphorase staining of samples of duodenum (Ward et al.,

1999). The density of Hu+ cells on the top surface of the

entire esophagus was determined in five Ret�/�, five Ret+/�,

Fig. 10. Low (A, B) and higher (C, D) magnification images showing the distrib

Ret+/� (A, C) and Ret� /� (B, D) mice. Many Phox2b+ cells were present along

number of Phox2b+ cells were present within the wall of esophagus of Ret� /� m

mice were in the rostral region. There were also some Phox2b+ cells in the stom

and five Ret+/+ mice. The density of Hu+ cells in the

esophagus of E18.5 Ret�/� mice was significantly lower

than that in the esophagus of both wild-type and Ret+/�

mice (ANOVA; Fig. 9), and was only about 4% that of wild-

type and heterozygous mice. The density of Hu+ cells in the

esophagus of Ret+/� mice was not significantly different

from wild-type mice (ANOVA; Fig. 9). A major sub-class of

neurons in the mouse esophagus expresses nitric oxide

synthase (NOS) (Sang and Young, 1997; Worl et al.,

2002). The proportion of Hu+ cells that showed NOS im-

munostaining was examined in three Ret�/�, three Ret+/�,

and three Ret+/+ mice. In wild-type and heterozygous mice,

around 40% of Hu+ cells were NOS+ (Fig. 9B). In contrast,

of 121 Hu+ cells examined in the esophagus of Ret�/� mice,

only 1 cell showed NOS immunostaining.

ution of Phox2b+ neural crest-derived cells within the esophagus of E11.5

the entire esophagus of the Ret+/� mouse (A, C). However, only a small

ice (B, D). Most Phox2b+ cells present in the esophagus of E11.5 Ret� / �

ach of Ret� / � mice (B). Scale bars: 50 Am.

H. Yan et al. / Developmental Biology 272 (2004) 118–133130

E11.5 esophagus

To determine whether the decrease in the density of

neurons in the esophagus of E18.5 mice is due to failure

of the neural crest cells to colonize the esophagus or a

failure to survive and differentiate, we examined the immu-

nolocalization of Phox2b in the E11.5 gastrointestinal tract.

Phox2b appears to be expressed by all neural crest-derived

cells within the gastrointestinal tract (Young et al., 2003). In

the esophagus of E11.5 wild-type (n = 7) and Ret+/� (n = 6)

embryos, many Phox2b+ cells were present along the whole

esophagus. However, in the esophagus of homozygous

mutant E11.5 embryos (n = 7), only sparse Phox2b+ cells

were seen within the wall of esophagus (Fig. 10). Most

Phox2b+ cells present in the esophagus of E11.5 Ret�/�

mice were in the rostral esophagus (data not shown). In

E11.5 homozygous mutant mice, Phox2b+ cells were ob-

served outside of the esophagus along the pathway of the

vagus nerve, and there were also Phox2b+ cells in the

stomach, but in vastly reduced numbers compared to wild-

type or Ret+/� embryos (Fig. 10).

Discussion

The results of this study show that (i) neurturin, like

GDNF, is chemoattractive to enteric neural crest-derived

cells and also induces neurite outgrowth; (ii) GDNF and

neurturin promote the migration of crest-derived cells from

the esophagus as well as the midgut and hindgut; (iii)

artemin has no detectable effect on migration or neurite

outgrowth in any region examined; (iv) the migratory

responses induced by GDNF or neurturin decline with

age; and (v) the development of esophageal neurons is

largely Ret-dependent.

Neurturin has similar effects to GDNF on migration and

neurite outgrowth

A previous study has shown that GDNF and neurturin

have similar effects on the survival, proliferation, and

differentiation of neural crest-derived cells isolated from

the gut (Taraviras et al., 1999). Our study showed that

exogenous neurturin also has similar effects to GDNF on the

migration of enteric crest-derived cells, and both neurturin

and GDNF were chemoattractive to crest-derived cells in

esophagus, mid-, and hindgut explants. The current study

also showed that both neurturin and GDNF induced neurite

outgrowth from explants of embryonic gut.

Neurturin is not essential for the migration of neural

precursors into and along the developing gut in vivo. Mice

lacking neurturin or its ligand binding molecule, GFRa2,

possess an enteric nervous system along the entire length of

the gastrointestinal tract (Gianino et al., 2003; Heuckeroth et

al., 1999; Rossi et al., 1999, 2003). Furthermore, although

there is some expression of neurturin in the mucosa of E12

mouse gut, neurturin expression is not detected in the gut

mesenchyme (through which neural crest cells migrate) until

E14 (Golden et al., 1999); this is after the colonization of the

embryonic mouse gut by crest-derived cells is complete

(Kapur et al., 1992). In contrast, GDNF signalling is

necessary for neural cell migration in the intestine (Cacalano

et al., 1998; Enomoto et al., 1998; Moore et al., 1996; Pichel

et al., 1996; Sanchez et al., 1996; Tomac et al., 2000), and

GDNF is expressed by the gut mesenchyme before the entry

of neural crest-derived cells (Natarajan et al., 2002). How-

ever, mice lacking neurturin or GFRa2, which forms part of

the receptor complex for neurturin, have fewer nerve fibres

in the circular muscle (Heuckeroth et al., 1999; Rossi et al.,

1999, 2003), where neurturin is normally primarily

expressed. Thus, neurturin seems to be important for induc-

ing axon outgrowth into the circular muscle layer.

Studies using cell lines have shown that neurturin can

activate Ret by binding to either GFRa1 or GFRa2 (Cree-

don et al., 1997; Jing et al., 1997). Hence, the responses

induced by neurturin described in the current study could be

due to neurturin binding to either GFRa1 or GFRa2, or to

both.

Neural cell migration and axon outgrowth have many

features in common (Rakic, 1999), and it is likely that the

response (migration or axon extension) induced by activa-

tion of Ret by GDNF or neurturin will depend on the age

and state of differentiation of the enteric crest-derived cells.

Ret activation in younger or undifferentiated enteric crest-

derived cells may bias the response toward migration,

whereas Ret activation in older, post-migratory cells may

bias the response toward neuronal differentiation and axon

outgrowth (Hearn et al., 1998; Wu et al., 1999).

Enteric neural crest-derived cells did not show a migratory

or neurite outgrowth response to artemin

Artemin appears to play a role in migration and neurite

outgrowth during the development of the sympathetic ner-

vous system (Enomoto et al., 2001; Honma et al., 2002;

Nishino et al., 1999). The only gastrointestinal region that

expresses artemin is the esophagus (Enomoto et al., 2001).

Despite the expression of artemin in the esophagus, enteric

neural crest-derived cells in both the esophagus and the

intestine failed to show a migratory or neurite outgrowth

response to artemin at the ages examined in this study.

However, we also showed that, unlike sympathetic neuron

precursors, enteric neural crest-derived cells do not appear

to express GFRa3, which would account for their lack of

response to artemin.

The migratory response induced by GDNF and neurturin

declines with age

Our data showed that the GDNF-induced migratory

responses in both the esophagus and midgut declined with

age because (a) transverse slices of E11.5 esophagus grown

on collagen gel showed a strong migratory response to

H. Yan et al. / Developmental Biology 272 (2004) 118–133 131

GDNF and neurturin, but E12.5 and E14.5 esophageal slices

showed little or no migratory response; and (b) the chemo-

attractive response to GDNF of crest-derived cells in sus-

pended explants of E12.5 midgut was dramatically lower

than that of E11.5 suspended midgut explants (see Fig. 4).

The earlier age-dependent decline in the esophagus com-

pared to the midgut parallels the earlier colonization of the

esophagus by neural crest cells. In addition, our study

showed that the GDNF-induced neurite outgrowth response

from the esophagus also declined with age. In fetal rats, the

proliferative responses and neuronal differentiation induced

by GDNF or neurturin also decline between E12.5 and

E14.5/E15.5 (Chalazonitis et al., 1998; Taraviras et al.,

1999).

The decline in responsiveness to GDNF does not appear

to be due to a down-regulation of GFRa1, as E14.5

esophageal neural cells showed strong GFRa1 immunos-

taining, but did not show GDNF-induced migratory or

neurite-outgrowth responses. Thus, the age-dependent de-

cline in responsiveness appears to be due to changes

downstream of GFRa1. It is unclear whether the decline

in the GDNF-induced migratory responses with age reflects

an age-dependent decrease in migratory ability, or other

factors such as an increase in cell–cell adhesion, which may

prevent cells from migrating.

Although GDNF induced a migratory response from

transverse slices of E11.5 esophagus grown on collagen

gels or on filter paper, we did not detect a migratory

response from suspended explants of E11.5 esophagus.

We believe this is because the crest-derived cells have to

migrate further to escape the tubular suspended explants

than they do to escape the transverse slices. Because the

esophagus of E10.5 mice is fragile, we could not reliably

set them up as suspended explants for quantitative

analysis.

There was also a decline with age of the migratory

response of neural cells in the esophagus induced by

neurturin. However, unlike GDNF, there was still a robust

neurite outgrowth response from the esophagus and mid-

gut induced by neurturin at E14.5, the oldest age exam-

ined. Because of the thickness of the gut wall, it was not

possible to establish suspension or slice cultures of gut

from older fetuses. However, as neurturin appears to be

involved in inducing axon outgrowth into the circular

muscle, and as the innervation of the circular muscle

continues to develop after birth (Young et al., 1998), it

appears likely that the neurites of enteric neurons will be

responsive to neurturin postnatally.

GDNF and neurturin promote the migration of crest-derived

cells from the esophagus as well as the mid- and hindgut

The rostrocaudal origin of neural crest cells that give rise

to esophageal neurons may vary between birds and rodent

(see Newgreen and Young, 2002). In mice, a DiI-tracing

study showed that esophageal neurons arise from neural

crest cells that emigrate adjacent to somites 6–7, whereas

intestinal neurons arise from neural crest cells that emigrate

from the hindbrain next to the more rostral somites (Durbec

et al., 1996). Neural crest cells adjacent to somites 6–7 also

give rise to dorsal root ganglia, and thus this level of the

neural axis can also be classified as ‘‘truncal’’ (Durbec et al.,

1996). In contrast, Burns et al. (2000) using cell labelling in

birds suggest that all esophageal neuron arise from the

rostral limit of the vagal region, not from the caudal or

truncal limit.

Because (a) esophageal neurons in embryonic mice arise

from a slightly different rostrocaudal level of the neural axis

from intestinal neurons (Durbec et al., 1996), (b) neurons

have been reported to be present in the esophagus and

stomach but absent elsewhere in the gastrointestinal tract in

Ret�/� mice (Durbec et al., 1996), and (c) neurons are absent

from the esophagus but present elsewhere in the gastrointes-

tinal tract of Mash1�/� mice (Guillemot et al., 1993), it

seemed likely that neural crest-derived cells in the esophagus

may have different trophic requirements and respond to

different signaling pathways from other enteric crest-derived

cells. However, the current study showed that, like intestinal

crest-derived cells, crest-derived cells in the esophagus

showed neurite outgrowth and migratory responses to both

neurturin and GDNF. Thus, at least in the assays performed in

this study, esophageal crest-derived cells behaved like other

enteric crest-derived cells, although the decline in the GDNF-

induced migratory response occurred in the esophagus before

more caudal regions of the gut.

Most esophageal neurons are dependent on Ret

The reported presence of neurons only in the esophagus

of Ret�/� mice, and the absence of neurons only from the

esophagus of Mash1�/� mice (Guillemot et al., 1993), has

lead to the idea that neural crest cells that migrate into the

esophagus have different trophic requirements from neural

crest cells that colonize the rest of the gut (Durbec et al.,

1996). However, Ret appears to be downstream of Mash1 in

signaling cascades (Lo et al., 1998), and if the development

of esophageal neurons were Mash1-dependent but Ret-

independent, it would suggest that there is a signaling

pathway downstream of Mash1, other than Ret, that is

essential for esophageal neuron development. However,

our study showed that there are very few neurons in the

esophagus of E18.5 Ret�/� mice and very few neural

precursors (Phox2b+ cells) in the esophagus of E11.5

Ret�/� mice. It therefore seems likely that few neural crest

cells ever migrate into the esophagus of embryonic Ret�/�

mice, and hence Ret signaling is required by essentially all

neural crest-derived cells destined to colonize the gut. The

small number of neural crest cells that reach the esophagus

and stomach probably only do so because they do not have

to migrate very far to reach these gut regions. Interestingly,

although gut neural crest stem cells are present in the

esophagus of Ret�/� mice, there are four times more stem

H. Yan et al. / Developmental Biology 272 (2004) 118–133132

cells in the esophagus of wild-type mice (Iwashita et al.,

2003).

There are different types of myenteric neurons in the

esophagus. In the current study, we showed that 40% of Hu+

cells expressed NOS in E18.5 heterozygote and wild-type

mice. However, of a total of 121 Hu+ neurons observed in

the esophagus of three Ret�/� mice, only 1 was NOS+. This

suggests either that the differentiation of all esophageal

neurons is delayed in the absence of Ret signaling, or that

Ret is specifically required for the development of NOS

neurons. In the developing esophagus, other major classes

of myenteric neurons (e.g., cholinergic neurons) do not

develop until after birth (Sang et al., 1999), so we were

unable to examine the appearance of other sub-populations

of neurons in Ret null mice.

The main difference between esophageal neurons and

other gastrointestinal neurons appears to be their dependen-

cy on Mash1; the development of all esophageal neurons is

Mash1-dependent, whereas only some sub-populations of

neurons in the intestine require Mash1 for their development

(Blaugrund et al., 1996). In addition to Mash1, Ret can also

be activated by Phox2b (Lo et al., 1998; Pattyn et al., 1999).

Phox2b�/� mice lack enteric neurons in all regions of the

gastrointestinal tract, including the esophagus (Pattyn et al.,

1999). It therefore seems likely that both Mash1 and Phox2b

are required for Ret activation in esophageal neuron pre-

cursors, but only Phox2b is required for Ret activation in

most intestinal neuron precursors.

Acknowledgments

This work was supported by the National Health and

Medical Research Council of Australia. We thank Drs. Jean-

Francois Brunet, Piers Emson, and Miles Epstein for kindly

providing antisera.

References

Airaksinen, M.S., Saarma, M., 2002. The gdnf family: signalling, biolog-

ical functions and therapeutic value. Nat. Rev. Neurosci. 3, 383–394.

Airaksinen, M.S., Titievsky, A., Saarma, M., 1999. GDNF family neuro-

trophic factor signaling: four masters, one servant? Mol. Cell. Neurosci.

13, 313–325.

Barlow, A., de Graaff, E., Pachnis, V., 2003. Enteric nervous system

progenitors are coordinately controlled by the G protein-coupled re-

ceptor EDNRB and the receptor tyrosine kinase RET. Neuron 40,

905–916.

Blaugrund, E., Pham, T.D., Tennyson, V.M., Lo, L., Sommer, L., Ander-

son, D.J., Gershon, M.D., 1996. Distinct subpopulations of enteric neu-

ronal progenitors defined by time of development, sympathoadrenal

lineage markers and Mash-1-dependence. Development 122, 309–320.

Burns, A.J., Le Douarin, N.M., 1998. The sacral neural crest contributes

neurons and glia to the post-umbilical gut: spatiotemporal analysis of

the development of the enteric nervous system. Development 125,

4335–4347.

Burns, A.J., Champeval, D., Le Douarin, N.M., 2000. Sacral neural crest

cells colonise aganglionic hindgut in vivo but fail to compensate for

lack of enteric ganglia. Dev. Biol. 219, 30–43.

Cacalano, G., Farinas, I., Wang, L.C., Hagler, K., Forgie, A., Moore, M.,

Armanini, M., Phillips, H., Ryan, A.M., Reichardt, L.F., Hynes, M.,

Davies, A., Rosenthal, A., 1998. GFRalpha1 is an essential receptor

component for GDNF in the developing nervous system and kidney.

Neuron 21, 53–62.

Chalazonitis, A., Rothman, T.P., Chen, J., Gershon, M.D., 1998. Age-

dependent differences in the effects of GDNF and NT-3 on the devel-

opment of neurons and glia from neural crest-derived precursors immu-

noselected from the fetal rat gut: expression of GFRalpha-1 in vitro and

in vivo. Dev. Biol. 204, 385–406.

Creedon, D.J., Tansey, M.G., Baloh, R.H., Osborne, P.A., Lampe, P.A.,

Fahrner, T.J., Heuckeroth, R.O., Milbrandt, J., Johnson Jr., E.M.,

1997. Neurturin shares receptors and signal transduction pathways with

glial cell line-derived neurotrophic factor in sympathetic neurons. Proc.

Natl. Acad. Sci. U. S. A. 94, 7018–7023.

Durbec, P.L., Larsson-Blomberg, L.B., Schuchardt, A., Costantini, F.,

Pachnis, V., 1996. Common origin and developmental dependence on

c-ret of subsets of enteric and sympathetic neuroblasts. Development

122, 349–358.

Enomoto, H., Araki, T., Jackman, A., Heuckeroth, R.O., Snider, W.D., John-

son Jr., E.M., Milbrandt, J., 1998. GFR alpha1-deficient mice have def-

icits in the enteric nervous system and kidneys. Neuron 21, 317–324.

Enomoto, H., Crawford, P.A., Gorodinsky, A., Heuckeroth, R.O., Johnson

Jr., E.M., Milbrandt, J., 2001. RET signaling is essential for migration,

axonal growth and axon guidance of developing sympathetic neurons.

Development 128, 3963–3974.

Fairman, C.L., Clagett-Dame, M., Lennon, V.A., Epstein, M.L., 1995.

Appearance of neurons in the developing chick gut. Dev. Dyn. 204,

192–201.

Focke, P.J., Schiltz, C.A., Jones, S.E., Watters, J.J., Epstein, M.L., 2001.

Enteric neuroblasts require the phosphatidylinositol 3-kinase pathway

for GDNF-stimulated proliferation. J. Neurobiol. 47, 306–317.

Focke, P.J., Swetlik, A.R., Schilz, J.L., Epstein, M.L., 2003. GDNF and

insulin cooperate to enhance the proliferation and differentiation of

enteric crest-derived cells. J. Neurobiol. 55, 151–164.

Gianino, S., Grider, J.R., Cresswell, J., Enomoto, H., Heuckeroth, R.O.,

2003. GDNF availability determines enteric neuron number by control-

ling precursor proliferation. Development 130, 2187–2198.

Golden, J.P., DeMaro, J.A., Osborne, P.A., Milbrandt, J., Johnson Jr., E.M.,

1999. Expression of neurturin, GDNF, GDNF family-receptor mRNA

in the developing and mature mouse. Exp. Neurol. 158, 504–528.

Golden, J.P., Milbrandt, J., Johnson Jr., E.M., 2003. Neurturin and perse-

phin promote the survival of embryonic basal forebrain cholinergic

neurons in vitro. Exp. Neurol. 184, 447–455.

Guillemot, F., Lo, L.C., Johnson, J.E., Auerbach, A., Anderson, D.J.,

Joyner, A.L., 1993. Mammalian achaete-scute homolog 1 is required

for the early development of olfactory and autonomic neurons. Cell

75, 463–476.

Hearn, C.J., Murphy, M., Newgreen, D., 1998. GDNF and ET-3 differen-

tially modulate the numbers of avian enteric neural crest cells and

enteric neurons in vitro. Dev. Biol. 197, 93–105.

Hearn, C.J., Young, H.M., Ciampoli, D., Lomax, A.E., Newgreen, D.,

1999. Catenary cultures of embryonic gastrointestinal tract support or-

gan morphogenesis, motility, neural crest cell migration, cell differen-

tiation. Dev. Dyn. 214, 239–247.

Heuckeroth, R.O., Lampe, P.A., Johnson, E.M., Milbrandt, J., 1998. Neur-

turin and GDNF promote proliferation and survival of enteric neuron

and glial progenitors in vitro. Dev. Biol. 200, 116–129.

Heuckeroth, R.O., Enomoto, H., Grider, J.R., Golden, J.P., Hanke, J.A.,

Jackman, A., Molliver, D.C., Bardgett, M.E., Snider, W.D., Johnson Jr.,

E.M., Milbrandt, J., 1999. Gene targeting reveals a critical role for

neurturin in the development and maintenance of enteric, sensory, para-

sympathetic neurons. Neuron 22, 253–263.

Honma, Y., Araki, T., Gianino, S., Bruce, A., Heuckeroth, R., Johnson, E.,

Milbrandt, J., 2002. Artemin is a vascular-derived neurotropic factor for

developing sympathetic neurons. Neuron 35, 267–282.

Iwashita, T., Kruger, G.M., Pardal, R., Kiel, M.J., Morrison, S.J., 2003.

H. Yan et al. / Developmental Biology 272 (2004) 118–133 133

Hirschsprung disease is linked to defects in neural crest stem cell func-

tion. Science 301, 972–976.

Jing, S., Yu, Y., Fang, M., Hu, Z., Holst, P.L., Boone, T., Delaney, J.,

Schultz, H., Zhou, R., Fox, G.M., 1997. GFRalpha-2 and GFRalpha-

3 are two new receptors for ligands of the GDNF family. J. Biol. Chem.

272, 33111–33117.

Kapur, R.P., Yost, C., Palmiter, R.D., 1992. A transgenic model for study-

ing development of the enteric nervous system in normal and agan-

glionic mice. Development 116, 167–175.

Kruger, G.M., Mosher, J.T., Tsai, Y.H., Yeager, K.J., Iwashita, T., Gariepy,

C.E., Morrison, S.J., 2003. Temporally distinct requirements for endo-

thelin receptor B in the generation and migration of gut neural crest

stem cells. Neuron 40, 917–929.

Le Douarin, N.M., Teillet, M.A., 1973. The migration of neural crest cells

to the wall of the digestive tract in avian embryo. J. Embryol. Exp.

Morphol. 30, 31–48.

Lo, L., Tiveron, M.C., Anderson, D.J., 1998. MASH1 activates expression

of the paired homeodomain transcription factor Phox2a, and couples

pan-neuronal and subtype-specific components of autonomic neuronal

identity. Development 125, 609–620.

Moore, M.W., Klein, R.D., Farinas, I., Sauer, H., Armanini, M., Phillips, H.,

Reichardt, L.F., Ryan, A.M., Carver-Moore, K., Rosenthal, A., 1996.

Renal and neuronal abnormalities in mice lacking GDNF. Nature 382,

76–79.

Natarajan, D., Marcos-Gutierrez, C., Pachnis, V., de Graaff, E., 2002. Re-

quirement of signalling by receptor tyrosine kinase RET for the directed

migration of enteric nervous system progenitor cells during mammalian

embryogenesis. Development 129, 5151–5160.

Newgreen, D., Young, H.M., 2002. Enteric nervous system: develop-

ment and developmental disturbances—Part 2. Pediatr. Dev. Pathol.

5, 329–349.

Nishino, J., Mochida, K., Ohfuji, Y., Shimazaki, T., Meno, C., Ohishi, S.,

Matsuda, Y., Fujii, H., Saijoh, Y., Hamada, H., 1999. GFR alpha3, a

component of the artemin receptor, is required for migration and sur-

vival of the superior cervical ganglion. Neuron 23, 725–736.

Norris, P.J., Charles, I.G., Scorer, C.A., Emson, P.C., 1995. Studies on the

localization and expression of nitric oxide synthase using histochemical

techniques. Histochem. J. 27, 745–756.

Pattyn, A., Morin, X., Cremer, H., Goridis, C., Brunet, J.F., 1997.

Expression and interactions of the two closely related homeobox

genes Phox2a and Phox2b during neurogenesis. Development 124,

4065–4075.

Pattyn, A., Morin, X., Cremer, H., Goridis, C., Brunet, J.F., 1999. The

homeobox gene Phox2b is essential for the development of autonomic

neural crest derivatives. Nature 399, 366–370.

Pichel, J.G., Shen, L., Sheng, H.Z., Granholm, A.C., Drago, J., Grinberg,

A., Lee, E.J., Huang, S.P., Saarma, M., Hoffer, B.J., Sariola, H., West-

phal, H., 1996. Defects in enteric innervation and kidney development

in mice lacking GDNF. Nature 382, 73–76.

Rakic, P., 1999. Neurobiology. Discriminating migrations. Nature 400,

315–316.

Rossi, J., Luukko, K., Poteryaev, D., Laurikainen, A., Sun, Y.F., Laakso,

T., Eerikainen, S., Tuominen, R., Lakso, M., Rauvala, H., Arumae,

U., Pasternack, M., Saarma, M., Airaksinen, M.S., 1999. Retarded

growth and deficits in the enteric and parasympathetic nervous sys-

tem in mice lacking GFR alpha2, a functional neurturin receptor.

Neuron 22, 243–252.

Rossi, J., Herzig, K.H., Voikar, V., Hiltunen, P.H., Segerstrale, M., Air-

aksinen, M.S., 2003. Alimentary tract innervation deficits and dys-

function in mice lacking GDNF family receptor alpha2. J. Clin.

Invest. 112, 707–716.

Sanchez, M.P., Silos-Santiago, I., Frisen, J., He, B., Lira, S.A., Barbacid,

M., 1996. Renal agenesis and the absence of enteric neurons in mice

lacking GDNF. Nature 382, 70–73.

Sang, Q., Young, H.M., 1997. Development of nicotinic receptor clusters

and innervation accompanying the change in muscle phenotype in the

mouse esophagus. J. Comp. Neurol. 386, 119–136.

Sang, Q., Ciampoli, D., Greferath, U., Sommer, L., Young, H.M., 1999.

Innervation of the esophagus in mice that lack MASH1. J. Comp.

Neurol. 408, 1–10.

Schuchardt, A., D’Agati, V., Larsson-Blomberg, L., Costantini, F., Pachnis,

V., 1994. Defects in the kidney and enteric nervous system of mice

lacking the tyrosine kinase receptor Ret. Nature 367, 380–383.

Shen, L., Pichel, J.G., Mayeli, T., Sariola, H., Lu, B., Westphal, H., 2002.

Gdnf haploinsufficiency causes Hirschsprung-like intestinal obstruction

and early-onset lethality in mice. Am. J. Hum. Genet. 70, 435–447.

Taraviras, S., Marcos-Gutierrez, C.V., Durbec, P., Jani, H., Grigoriou, M.,

Sukumaran, M., Wang, L.C., Hynes, M., Raisman, G., Pachnis, V.,

1999. Signalling by the RET receptor tyrosine kinase and its role in

the development of the mammalian enteric nervous system. Develop-

ment 126, 2785–2797.

Tomac, A.C., Grinberg, A., Huang, S.P., Nosrat, C., Wang, Y., Borlongan,

C., Lin, S.Z., Chiang, Y.H., Olson, L., Westphal, H., Hoffer, B.J., 2000.

Glial cell line-derived neurotrophic factor receptor alpha1 availability

regulates glial cell line-derived neurotrophic factor signaling: evidence

from mice carrying one or two mutated alleles. Neuroscience 95,

1011–1023.

Ward, S.M., Ordog, T., Bayguinov, J.R., Horowitz, B., Epperson, A., Shen,

L., Westphal, H., Sanders, K.M., 1999. Development of interstitial cells

of Cajal and pacemaking in mice lacking enteric nerves. Gastroentero-

logy 117, 584–594.

Widenfalk, J., Nosrat, C., Tomac, A., Westphal, H., Hoffer, B., Olsen,

L., 1997. Neuturin and glial cell line-derived neurotrophic factor

receptor-h, novel proteins related to GDNF and GDNFR-a with

specific cellular patterns of expression suggesting roles in the develo-

ping and adult nervous system and in peripheral organs. J. Neurosci.

17, 8506–8519.

Worl, J., Dutsch, F., Neuhuber, W.L., 2002. Development of neuromuscular

junctions in the mouse esophagus: focus on establishment and reduction

of enteric co-innervation. Anat. Embryol. (Berl.) 205, 141–152.

Worley, D.S., Pisano, J.M., Choi, E.D., Walus, L., Hession, C.A., Cate,

R.L., Sanicola, M., Birren, S.J., 2000. Developmental regulation of

GDNF response and receptor expression in the enteric nervous sys-

tem. Development 127, 4383–4393.

Wu, J.J., Chen, J.X., Rothman, T.P., Gershon, M.D., 1999. Inhibition of in

vitro enteric neuronal development by endothelin-3: mediation by endo-

thelin B receptors. Development 126, 1161–1173.

Xian, C.J., Huang, B.R., Zhou, X.F., 1999. Distribution of neurturin

mRNA and immunoreactivity in the peripheral tissues of adult rats.

Brain Res. 835, 247–258.

Yntema, C.L., Hammond, W.S., 1954. The origin of intrinsic ganglia of

trunk viscera from vagal neural crest in the chick embryo. J. Comp.

Neurol. 101, 515–541.

Young, H.M., Newgreen, D., 2001. Enteric neural crest-derived cells: or-

igin, identification, migration, differentiation. Anat. Rec. 262, 1–15.

Young, H.M., Torihashi, S., Ciampoli, D., Sanders, K.M., 1998. Identifi-

cation of neurons that express stem cell factor in the mouse small

intestine. Gastroenterology 115, 898–908.

Young, H.M., Hearn, C.J., Farlie, P.G., Canty, A.J., Thomas, P.Q., New-

green, D.F., 2001. GDNF is a chemoattractant for enteric neural cells.

Dev. Biol. 229, 503–516.

Young, H.M., Bergner, A.J., Muller, T., 2003. Acquisition of neuronal

and glial markers by neural crest-derived cells in the mouse intestine.

J. Comp. Neurol. 456, 1–11.